The Third Glyph
A deep dive into skill combos in Shroud of the Avatar's deck combat, what thirty years of combining mechanics can teach it, and eighty concrete proposals, including a compendium of fifty-eight new combos.
Somewhere in the middle of a fight, a player drags one glyph onto another and gets a third skill that exists nowhere on any tree: Shield Bash and Flame Fist become Flaming Bash. It is the single most distinctive thing SotA's deck combat does, the mechanical answer to why skills are cards at all. The machinery behind that moment is in better shape than almost anyone knows: three ways to combine, live highlighting of valid pairs, a full combo list with icons and tooltips, experience that flows back into both glyph schools. And almost nobody knows, because the game never says a word about any of it. Twenty-nine combos serve twenty-one schools; six schools have no combo at all; the Chaos school, whose whole identity is transmutation, participates mostly through a single spell. This document tours the system as it actually works, measures it against every game that ever let players combine abilities, proposes twenty-two system improvements, and then does the thing players have asked for since 2016: it sits down and writes them: fifty-eight new combos, every school included, sixteen of them for Chaos.
The combo has three birthplaces. It was born as a kept bug in Street Fighter II in 1991, where canceling one attack into another became the genre's definition of skill. It grew a soul in Chrono Trigger in 1995, where two characters' techniques fused into a third with its own name and its own fanfare. And it learned to live in a persistent world in Final Fantasy XI in 2002, where two weapon skills landed in rhythm detonated a named chain in the party's log. Ultima Online, the game SotA descends from, had none of this; deck combat and glyph fusion are SotA's own addition to the inheritance. That addition is real: a fusion verb performed live under pressure, named results, levels that follow your training, experience that flows back into both schools. Measured against every game that ever combined abilities, SotA's machine is ahead of most of them. What it never built is the part every one of those games is remembered for: the moment the game tells you.
The gaps cluster in three places:
What follows is eighty proposals in two very different sections. Section 3 is the system: twenty-two cards, nine of them small client fixes, that make the table true, tell the players, and then deepen the machine, ending at three genuine design forks that belong on the owner's desk. Section 4 is the content: a compendium of fifty-eight fully specified new combos, every adventuring school included, the six silent schools given voices, sixteen combos for Chaos alone, each one collision-checked against every pairing that exists or has been promised anywhere. Every card and every combo on this page carries its own vote and its own discussion thread.
| # | Proposal | Tier | Size | One-line pitch |
|---|---|---|---|---|
| S1 | Wake the lost combo tutorial | 1 | S | the finished first-combo popup finally fires |
| S2 | An honest Combo Glyph label Built | 1 | S | tooltips stop lying about what combines |
| S3 | Safe defaults for the stack math Built | 1 | S | a bad fallback can no longer break focus costs |
| S4 | Lint the combo table Built | 1 | S | bad entries get caught the day they are written |
| S5 | A school filter that answers the question | 1 | S | filter to Fire, see Fire's combos |
| S6 | A combo list you can open anywhere | 1 | S-M | the list moves to where players actually look |
| S7 | A combo journal with fanfare | 2 | M | first discoveries celebrated and kept forever |
| S8 | Say the combo's name out loud | 2 | S | every combo announces itself, like skillchains did |
| S9 | Every glyph tells you its combos Built | 2 | S-M | Fire Arrow's tooltip finally mentions Fiery Decay |
| S10 | Combinable glyphs shimmer | 2 | S | the hand shows what can fuse, as a setting |
| S11 | Trainers, tomes, and a combo quest | 2 | M-L | combos taught in the world, never gated by it |
| S12 | Achievement breadcrumbs | 2 | S | the journal whispers that combos exist |
| S13 | A safe first swing | 2 | M | one practice moment at a trainer's dummy |
| S14 | Unlock alternate pairings | 3 | M | one combo, several ways to reach it |
| S15 | Count what players combine | 3 | M | a usage board ends the guessing |
| S16 | Rebalance the stack math | 3 | M | the five dials get a deliberate pass |
| S17 | Combos for locked decks | 3 | M | fused glyphs land as overlays on locked slots |
| S18 | Higher order combos | 3 | XL | a handful of three-part showpieces, owner willing |
| S19 | Statuses set the table | 3 | L | a frost combo hits harder on a burning target |
| S20 | Should combos grow on their own? | 4 | fork | derived levels, or a life of their own |
| S21 | Trainable, lockable combos? | 4 | fork | is the combo the act, or the possession |
| S22 | Show every combo, or make them found? | 4 | fork | full listing versus silhouettes, ruled before S7 |
| — | The combo compendium | 4 | — | fifty-eight new combos, every school at the table, votable one by one |
Credit first, because the verb itself is genuinely well built. There are three ways to make a combo, all live right now: drag one glyph onto another in the combat hand; press and hold one glyph's key and then press a second glyph's key; or press R to enter combo mode, which puts a banner on the hotbar and lets you pick the two glyphs at leisure. While you decide, the hand quietly tests every pairing in the background and puts a glow on any two glyphs that can actually combine, and flashes the slot red if you try a pair that cannot. Stacking two copies of the same glyph and fusing two different ones ride the same gesture. The result lands in the second glyph's slot with a full fresh timer, and casting it trains both glyph schools at once. A combo glyph even reads its level honestly: the average of its two parents, rounded up, and zero if you have not learned both.
There is even a combo list. The Deck Building window has a Combos tab that lists all twenty-nine combos with the result's icon, both glyphs' icons, tooltips for each part, a school filter, and a toggle to grey out or hide the ones you cannot make yet. And there is a finished help window in the game files that begins, in its own words, by telling the player they have just triggered the ability to perform a Combo Glyph. It explains the drag, the R key, everything. Nothing in the game ever opens it. That sentence is this page in miniature.
| Combo | Made from |
|---|---|
| Flaming Bash | Shield Bash + Flame Fist |
| Fiery Decay | Fire Arrow + Death Ray |
| Titan Smash | Knockback + Body Slam |
| Sacrifice | Healing Ray + Death Ray |
Now the ledger's other column. Here is the entire journey of a new player meeting this system, annotated:
Every card added to a glyph, by stacking or combining, moves five dials at once. These are the live values:
| Dial | Per extra card | Note |
|---|---|---|
| Damage | +20% | damage effects only |
| Other effect strength | +5% | buffs, debuffs, heals |
| Effect duration | +25% | longer burns, longer buffs |
| Cooldown | −0.5s | faster to use again |
| Focus cost | discounted | compounds with each card |
A fused combo starts at the lower of its two parents' card counts, caps at six like everything else, and pays one price a stack never pays: both glyphs go on cooldown when it fires. The five dials are generic; no dial exists that a specific combo can turn. Those are design choices worth revisiting deliberately rather than by archaeology, and the Tier 3 cards do.
For the technically curious, the whole pipeline is data driven and lives in the deck code. Each combo is authored on the result: the combo glyph's data names the two glyphs that make it, and at load time the game builds a lookup table registering both orders. The combining gesture asks that table what a pair makes, and the answer is final; pairings are facts of the data, not chances.
flowchart LR
D[Deck draws glyphs] --> H[Combat hand]
H -->|drag A onto B| C{Combine}
H -->|hold A, press B| C
H -->|R, then A and B| C
C -->|same glyph| S[Stacked glyph
higher card count]
C -->|combo pair| F[Combo glyph
in B's slot]
C -->|no combo| X[Red flash]
F --> U[Cast]
U --> XP[XP flows to both
glyph schools]
The idea that two abilities can make a third has three separate birthplaces, and SotA inherited from all of them. The fighting-game root is Street Fighter II in 1991, where the combo was literally a bug: an input buffer more forgiving than intended let players cancel one attack into another, and the designers decided the accident was a skill worth keeping. That is the lineage of SotA's real-time gesture, the drag under pressure. The party-RPG root is Chrono Trigger in 1995, whose Dual Techs fused two characters' known techniques into a brand new named ability with its own animation and its own fanfare the moment it was learned. That is the lineage of the named combo glyph. And the MMO root is Final Fantasy XI in 2002, whose skillchains detonated a named event in the combat log when two compatible weapon skills landed in sequence, opening a window other players could burst into. That is the lineage nobody has claimed yet in SotA, and section 3 has opinions about it.
One lineage is conspicuously absent. Ultima Online never combined anything: its magic was a fixed spellbook behind reagent words, and its combat was timing and gear. Deck combat and glyph fusion are things SotA added on top of its inheritance, not things it carried forward. There is no way Ultima did it to appeal to. That cuts both ways: the system owes no debts, and it has no map. The games below are the map.
Final Fantasy XI built the MMO's first composition system and then taught nobody: skillchain compatibility lives on a hidden attribute graph that the community reverse-engineered into charts it has now maintained for over twenty years, and knowing the chart became a badge that pick-up groups were judged by. What survived is the part everyone remembers fondly: the named detonation in the log, witnessed by the whole party, that everyone contributed to. FFXIV dropped the system and its forums still ask for it back. The lesson: announce the combo by name where everyone can see it, and never leave the chart to the wiki.
Chrono Trigger is still the gold standard for telling the player. The moment both prerequisites of a Dual Tech were met, the game said so, and the fused tech sat in the menu from then on with both its techs visible. Discovery felt like a reward even though it was automatic. Its other lesson is quieter and cost Magus dearly: fused abilities are content, content has a quadratic authoring bill across characters, and endgame players skip fusions that lose to two strong singles. A combo has to beat the sum of its parts somewhere, or it is decoration.
Guild Wars 2 is the cautionary tale this page returns to most. Fields and finishers are tagged on every tooltip, cross-player, and mechanically lovely, and the official forums are full of veterans discovering after hundreds of hours that the system exists. Documented is not taught. Its second failure is subtler: when overlapping fields resolve by oldest-wins, your heal blast hits a stranger's stale fire field, and when the payoffs decay to a sliver of Might, the system quietly becomes flavor. Outputs must stay strong and resolution must be predictable.
Magicka made discovery itself the game, twice. Its element mixing teaches through spectacular visible failure, which an MMO cannot copy wholesale, but its authored layer can be copied exactly: named signature spells are learned by finding their tomes in the world, each one a small event. Spells as loot turns exploration into combat vocabulary. And even Magicka, the most dexterity-proud combining game ever made, lets you hotkey a learned Magick. There is always a lower-dexterity path to the thing you already earned.
Divinity: Original Sin 2 proved a combo grammar can teach itself when the state is visible and the logic matches intuition: everyone alive already knows oil burns and water conducts lightning. Nobody reads a Divinity combo list, because the world is the list. Its warning is just as famous: the armor system that gated every status payoff made the elegant grammar read as broken, and late fights drowned the whole layer in surface soup. Visible state, ungated payoff, legible scale.
The Elder Scrolls Online solved teaching outright. A synergy puts a prompt on an ally's screen at the exact moment it is usable, names the key, and costs nothing to accept. One hundred percent of ESO players know synergies exist; there is no tutorial, because the mechanic is its own. The price is that a free prompt carries no skill expression at all. SotA, which already has a real dexterity verb, is in the rare position of being able to add the prompt without giving up the craft.
Path of Exile composes differently: support gems do not merge two abilities, they modify one, and the whole composition happens out of combat in the build layer. It is the proof that modification multiplies variety combinatorially where bespoke fusion is priced linearly, and its own sequel is the proof of the other lesson: PoE 2 deleted the socket RNG that stood between players and composition. Whatever composition SotA adds, acquiring the parts must be deterministic.
Briefer mentions that earn their line: Outward paces combining as prepare-then-trigger, a sigil on the ground transformed by the next cast into it, deliberate rather than twitchy. Lost Ark shows constrained modification, pick one of three per tier, reads instantly where PoE's open graph overwhelms. And the fighting-game lineage spent thirty years building what every dexterity combo system owes its players: a practice room, a trial list, and feedback on the attempt.
Rows are capabilities a combining system can have. Filled means the game has it well, half means partially or poorly, empty means absent. The two right columns are this page's argument. Read the SotA column honestly before the gaps: on first-class named fusion and on results that scale with your investment, SotA is ahead of every game on this table. The system's problem was never the machine.
| Capability | FFXI | CT | GW2 | Magicka | DOS2 | ESO | SotA today | SotA proposed |
|---|---|---|---|---|---|---|---|---|
| Named fusion: two abilities make a first-class third | ◑ | ● | ○ | ● | ○ | ○ | ● | ● |
| In-game combo reference with its glyphs | ○ | ● | ○ | ◑ | ○ | ○ | ◑ | ● |
| Prompt or hint at the actionable moment | ○ | ○ | ○ | ○ | ◑ | ● | ◑ | ● |
| First-time fanfare: discovery as a reward | ○ | ● | ○ | ◑ | ○ | ○ | ○ | ● |
| Named announcement on execution | ● | ◑ | ○ | ○ | ○ | ○ | ○ | ● |
| In-world teaching: trainers, tomes, quests | ○ | ◑ | ○ | ● | ○ | ○ | ○ | ● |
| Safe practice space for the gesture | ○ | ○ | ◑ | ○ | ○ | ○ | ○ | ● |
| Collection and achievement tracking | ○ | ○ | ◑ | ○ | ○ | ◑ | ○ | ● |
| Depth beyond a single pair | ● | ● | ◑ | ● | ◑ | ○ | ◑ | ◑ |
| Alternate pairings for one result | ● | ○ | ● | ◑ | ● | ○ | ○ | ● |
| Status and environment interplay | ◑ | ○ | ◑ | ● | ● | ○ | ○ | ◑ |
| Low-dexterity path to the same result | ○ | ● | ○ | ◑ | ● | ● | ◑ | ● |
| Payoff scales with your investment | ◑ | ◑ | ○ | ○ | ○ | ○ | ● | ● |
| Cross-player composition | ● | ● | ● | ◑ | ◑ | ● | ○ | ○ |
The last row stays empty in both SotA columns on purpose: the Party and Grouping page weighed cross-player combo abilities and set them aside, and this page honors that ruling rather than reopening it.
Twenty-two proposals about the system itself, in three working tiers and an owner's desk. They deliberately mint no new combos; the combos get their own section, and it is bigger than this one. Sibling school pages already own several combo cards, and this page credits rather than repeats them: the Tactics page owns the melee combo mechanism its ring is waiting on, the Blades page proposed a forms panel and a starter-deck moment, the Bludgeons page proposed tooltip readback for its own school. Where a card below generalizes one of those, it says so.
The game already contains a finished help popup that congratulates you the first time a combo becomes possible, explains the drag, the hold, and the R key, and points at the combo list. It has full text and a window design, and no player has ever seen it, because the moment meant to show it was never wired up. Fire it the first time two glyphs in your hand can combine, once per character, respecting the usual help toggles.
Seen in: Elder Scrolls Online (teach at the exact actionable moment); Chrono Trigger (discovery as a designed moment).
The orphaned content is HelpContext.Combos with its window asset and text in the help data files. The hand already dry-runs every pair each frame to drive the combinable glow (CombinableSlotsManager in CombatHandUI); the first frame any pair reports combinable is the trigger. Persist the seen-flag alongside the other first-time help flags. Keep it non-modal; if mid-fight timing tests poorly, defer the window to combat end and show only a toast in the moment.
Built August 15. Seven glyphs advertised a combination no pairing could produce, and the four elemental summons could be made without ever saying so. Both directions now match the combo table, and an editor test compares the two on every run. Not on the test shard yet.
Some skills say Combo Glyph in their tooltip when no pairing can make them, and the four elemental summons that really are combos do not say it at all. Make the label agree with the combo table exactly, in both directions, so the one place the game does whisper about combos stops whispering wrong.
Seen in: nowhere; this is housekeeping.
The label came from a hand-maintained skill group list (Skill.GetGroupName) that had drifted from the data in both directions. Seven entries were tagged as combos with nothing that could make them: Earth Shatter, Hamstring (Dexterity) and the two Summon Ankh results have rune data with an empty pairing line, and Pulverize, Piercing Flurry and Chaotic Purge have no rune data at all. Those moved to a separate group, which drops the claim from their tooltip and leaves them where the compendium and the Earth page can revive them. The four elemental summons could not be fixed in that list at all, because the group lookup returns early for anything that belongs to a school, so a Combo Glyph entry for Summon Air Elemental would have been unreachable code; the skill type now asks the combo table before it asks the school. An editor test asserts the group holds exactly the school-less glyphs a pairing can produce, and fails the day one of the parked seven is given a pairing.
Built August 15. Four of the six combo tuning dials disagreed with the values the game actually ships, including the one that would have turned the focus cost of every stacked glyph into an invalid number. All four now match, the formula clamps its input, and an editor test holds the code defaults and the shipped tuning file together. Not on the test shard yet.
One of the combo focus-cost values has a code-side default that produces an invalid number, and only the tuning file's override keeps it out of the game. Set the defaults to the live values and clamp the formula so the math cannot go bad.
Seen in: nowhere; this is housekeeping.
The focus discount is an exponential curve in the card count, and the code-side default for its rate put a negative number under a fractional exponent, which is not a number. The clamp at the end of the formula could not catch it, because the larger of zero and an invalid number is the invalid number, so it reached the focus check and every stacked glyph would have read as unaffordable. The route to a live break is the tuning file itself: it leaves any key it does not find at whatever the code declared, which has already happened once to another field. Four of the six combo dials disagreed with the shipped values, including the maximum card count; all four were corrected, and the rate is clamped before it reaches the curve.
Built August 15. Nine checks now run over every combo whenever the rune data is validated, covering a pairing with a glyph missing, a glyph that can never reach the combining step, a passive standing in for a glyph, and a pair claimed by two different combos. All twenty-nine combos in the game pass, so nothing changed for players today; the point is that the fifty-eight below cannot land broken. Not on the test shard yet.
The combo data has no automatic checks: a pairing missing a glyph, a passive listed where an active belongs, or two combos claiming the same pair would only be caught when a player noticed something wrong. Add validation that runs when the data is built, so a bad entry is caught the day it is authored. With fifty new combos proposed below, the table deserves a linter before it grows.
Seen in: every card game's database checks.
The rules live in their own file as a plain function over a flattened view of the table, which is what makes them testable without fabricating rune data. The rune validation builds that view from the same asset group and skill table it already loads, so the checks join the existing Validate menu and the build server's validation step with no new wiring, and they deliberately do not touch the runtime lookup tables the game fills at load. Nine rules: an empty entry, a glyph reference not set, a glyph with no id, a glyph that is the discard-timer rune, a glyph that is the combo itself, a glyph missing from the shipped rune set, a glyph with no skill row, a glyph the skill table does not consider castable, and one pair claimed twice. Four of those are silent today, because the combining step turns them away before it ever consults the table.
The lookup tables were also append-only, so validating twice in one session reported every combo as a duplicate of itself; they now rebuild on each load, and the duplicate warning names both combos that claimed the pair instead of one of the glyphs. The suite is twenty-one editor tests: one per rule, plus a pin that the linter really is reading the shipped table and one that a combo stealing an existing pair is caught. Owner call taken: a bad entry fails validation rather than the client build, so nobody's build stops mid-flight over a data typo.
The combo list's school filter currently hides a combo unless both glyphs' schools are ticked. Filter to just Fire, the reasonable thing for a fire mage to do, and most Fire combos vanish, because their partner school is unticked. Change the filter so ticking a school shows every combo that school takes part in.
Seen in: standard filter behavior everywhere.
The row visibility check requires each glyph's school to pass the filter; one condition flips from and to or. The saved filter preference keeps its key and shape.
A complete combo list already exists, with pairings, glyph icons, tooltips, and greyed-out entries for what you cannot make yet. It lives only inside deck building, so a player who never edits a deck never meets it. Give the same list a second home reachable any time: an entry from the skills window and a chat command. Later cards turn it into the combo journal; this one just moves it where people look.
Seen in: Chrono Trigger (techs listed in the main menu); Magicka (the spellbook).
Pure reuse of the existing combo list control and its dialog wrapper; the work is window plumbing and an entry point in the skills window. Ships after S2 and S5 so the list it surfaces is truthful and filterable.
This tier is the answer to the question this page exists for. The genre's verdict is unambiguous: a reference list alone is the documented-but-not-taught failure, and the fix is one push channel that volunteers the information during play plus one world channel that makes discovery into content. This tier ships both, twice over.
The first time you pull off each combo, the game should say so: a short flourish, the combo's name on screen, and a permanent entry in a personal combo journal recording the pairing. Discovery becomes a reward you keep. The journal is the combo list from S6 grown up: your discovered combos shine, the rest wait however the S22 ruling decides.
Seen in: Chrono Trigger (new techs announce themselves and stay listed with the techs that make them); Magicka (a spellbook that fills as you find Magicks).
Hook the successful fuse in the deck controller; discovered state is a small per-character record using fixed numeric identifiers, a modest addition to the character data the servers already persist. Toast and sound are client work. The journal view is the S6 list plus a discovered flag per row.
Every combo already has a name, and today performing one prints nothing. Announce it: a combat log line and an optional overhead flourish when a combo fires, visible to the people fighting beside you. Twenty years ago a generation of players learned that skillchains existed because the log kept saying so, by name. Every execution becomes an advertisement.
Seen in: Final Fantasy XI (named skillchains in the party log).
The combo's identity is known wherever its effect is shown; add a log channel line and floating text at apply time. Options toggle with a spam-conscious default: your own combos announce, others' are opt-in.
Built August 15. Every glyph that is half of a pairing now says so in its tooltip, everywhere tooltips appear: the skill tree, the skills window, the deck builder, and the hand in combat. Each line names the combo and the partner glyph it needs, and partners you have not learned show in red. A glyph in many combos shows six lines and a count of the rest, pointing at the Combos tab. The lookup reads the full pairing-set list from day one, so alternate pairings will need no rework here. Not on the test shard yet.
No skill tooltip anywhere mentions that the skill is half of a combo. Add a breadcrumb to every glyph's tooltip listing the combos it takes part in and the partner needed, everywhere tooltips appear: skill tree, trainer, deck builder, hand. A player studying Fire Arrow learns about Fiery Decay at exactly the moment they are deciding whether Fire Arrow is worth training. The Bludgeons page proposed this for its own school; this is the system-wide version, and absorbs it.
Seen in: Guild Wars 2 (tooltip tagging, together with its lesson that tooltips alone are not enough, which is why this tier has six more cards).
Build a reverse glyph-to-combos index from the combine lookup at load; append a section in the shared skill tooltip formatter. Written against the full pairing-set array from day one so S14 needs no rework. Long lists cap with a pointer to the journal; one Chaos glyph currently sits in nine combos.
The game already computes, every moment of combat, which glyphs in your hand can combine; it only shows you inside combo mode. Promote that knowledge to an always-on shimmer on combinable pairs, controlled by a settings toggle, subtle enough that veterans keep it and obvious enough that a newcomer's eye catches it. The Blades page proposed a shimmer for its own forms; this is the one system setting they all ride.
Seen in: Elder Scrolls Online (the prompt at the moment is the whole tutorial), translated to a hand of cards.
The per-frame pair dry-run already pays the compute cost for combo mode's highlighting; this adds a persistent visual state on hand slots outside combo mode plus an options entry. Default on for new characters, and the S1 popup points at the setting.
Combos should be taught inside the world, not only by the interface. Combat trainers get a dialogue topic that explains combining and names a combo or two from their school. Collectible combo tomes, found as loot and sold by the odd vendor, reveal a combo in your journal when read. One short quest walks a new player through their first combo and rewards a tome. Nothing is locked behind any of this: every combo keeps working for everyone, taught or not. Teach, never gate.
Seen in: Magicka (Magicks learned from tomes found in the world); Outward (combat trainers who sell the knowledge).
Trainer topics ride the conversation pipeline; a tome is an item whose use marks a journal entry revealed and opens the journal to it; the quest reuses the S13 practice moment. Content and localization are the real cost, so the first pass scopes to one trainer topic sweep and a starter tome set.
The achievements journal should whisper that combos exist: perform your first combo, perform ten different combos, stack a glyph to six cards, combine across two schools. Each is a nudge toward the combo journal for the players who read achievement lists, at almost no cost. Support pattern, not a solution; it ships alongside the rest of the tier, never instead of it.
Seen in: every modern collection game; the research's verdict is cheap and never sufficient alone.
Combo identity and stack depth are known at use time; increment fixed-identifier stats where skill-use stats already increment, and hang achievement rules on them through the existing pipeline.
Combining is a thing you do under pressure, and pressure is a bad classroom. Give players one safe moment to try it: a sparring session at a combat trainer's dummy where the trainer deals you two combinable glyphs and asks you to put them together, with the gesture shown on screen. Thirty years of fighting games agree that a combo system owes its players a practice room. One moment, not a tutorial rework. The Blades page proposed a starter-deck first-combo beat; it can be this one.
Seen in: fighting game trials; Elder Scrolls Online (the prompted moment).
Training dummies exist in scenes; the session is a scripted prompt sequence over a dealt hand with a known combinable pair, and success fires the S7 fanfare, which doubles as the quest beat for S11. The S1 popup remains the zero-cost fallback for anyone who skips it.
The data format already allows a combo to have more than one pairing, and the deck builder even has an or display ready for it. But every piece of code that reads them assumes exactly one, so the feature has never been usable: authoring a second pairing today would quietly break the list and tooltips. Fix the readers, and one combo can be reachable from two or three glyph pairs. That multiplies how many real player builds can reach a result without minting a new result for every pair.
Seen in: Final Fantasy XI (many weapon pairings produce the same named chain); Guild Wars 2 (any finisher of a type completes any field of a type).
Replace the first-pairing-only assumptions in the combo list, availability check, and attunement math with all-sets readers; a displayed level becomes the best learnable path while a fused glyph levels off the pair actually consumed. The existing combo list tests extend to pin the or rendering. Lands before the compendium slate ships so new combos can use it from day one.
Nobody can currently say which combos players actually perform. The skill-use reporting already carries whether a use was a combo and which glyphs fed it; index those fields and build a combo usage board: most performed, never performed, typical stack depth. Every balance question in this system becomes answerable, and the never-performed list becomes the pruning and teaching worklist. The recently shipped public leaderboards can carry a combos board when there is something worth showing.
Seen in: live-ops practice everywhere; this page's own school-page siblings lean on exactly this data for single skills.
Combo identity already travels with each use report; the work is extending the indexed skill-use record with the combo fields and a small dashboard view. A recent fix (#4543) corrected how combo identity was read, so the trustworthy-data clock starts after it; the retune card below shares the same thirty-day wait the Ranged page committed to.
Every combo in the game shares one bonus table per card: damage up twenty percent, other effects up five, duration up twenty-five, cooldown down half a second, a compounding focus discount, cap at six. Damage grows four times faster than everything else, which quietly tells players to only stack damage glyphs. And a fused combo waits on both of its glyphs' cooldowns while a stack waits on one, which taxes exactly the players who built a deck around a favorite combo. This card is the deliberate tuning pass over those dials, driven by the S15 data, changed in tuning data only.
Seen in: Guild Wars 2, as the warning: combo payoffs that decay into flavor take the whole system with them.
All five dials live in the global tuning asset. The candidate knobs, in order of suspicion: the damage-versus-effect growth gap, the both-cooldowns rule for fused results, the focus curve, and the six-card cap. Ships through the balance changeset pipeline with the usual preview discipline, only after thirty days of clean telemetry, coordinated with the Ranged page's own combo retune.
A locked glyph can start a combo but can never receive one, so heavily locked decks are shut out of the system they might enjoy most. Let a combo result land on a locked slot as a temporary overlay: the fused glyph sits on top until used or expired, then the locked glyph underneath returns. Locked decks keep their tradeoff and gain the verb.
Seen in: nowhere directly; this answers a standing player complaint the Earth page recorded.
The target-slot gate is one check in the deck controller; the overlay is a slot state whose lifetime maps to the fresh discard timer a fused result already receives. The locked-glyph-as-source behavior, including its cooldown rule, stays exactly as it is.
A combo result plus a third compatible glyph could make a small set of grand finishers. The precedent says scarcity is the whole design: the game that invented chain tiers shipped exactly two top-tier chains, and that is why landing one was an event. If this is ever built it is a handful of authored showpieces for the dexterity elite, not a second combo table. Parked on the owner's desk until the pair layer is proven healthy.
Seen in: Final Fantasy XI (level two and three skillchains); Chrono Trigger (triple techs).
Mechanically a fused result has an identity and can in principle feed a pairing; the real work is design: card counts for three-way results, hand readability, and the authoring bill the Chrono Trigger warning is about. Waits on S14 and S15 at minimum.
Some combos should care what state the target is in. A frost combo against a burning target, a shock combo against a soaked one: the same pairing, an upgraded result when the fight has been set up right. Target statuses are already visible on screen, so this layer teaches itself through play with no new interface at all. Of the sequencing mechanics the genre offers, this page picks exactly one, and this is it: upgrades touch damage and duration only, never crowd control, and nothing is gated in front of the payoff.
Seen in: Divinity Original Sin 2 (a visible-state grammar players learn by intuition); Magicka (element interactions).
Authored as conditional effect blocks on the result runes' existing effect data, keyed on target status; no change to the combine path or interface. Pilot on a handful of elemental combos, watched through S15. Chosen over burst windows and primer glyphs, which would each need new interface and timing machinery; the considered-and-rejected list has the full reasoning. The one anti-pattern honored: no gate in front of the upgrade, which is the mistake that soured Divinity's own system.
Today a combo's level is the average of its two glyphs, rounded up, and it never earns experience itself; casting it trains those glyphs. Three roads: keep that, and mastering the parts is mastering the whole; give combos their own experience and a place in progression; or a middle path where a combo proficiency tunes only its stack bonuses. Each changes what training means, so it is presented as a fork, not a recommendation.
Seen in: Chrono Trigger (the parts rule) versus Path of Exile (the composed thing levels on its own).
The experience split and the derived-level rule are each one well-contained routine today. The structural blocker for independent progression is that combos live outside every school tree, which is also what keeps them free; any road that changes it is priced accordingly.
Combos cannot be bought, trained, or locked into a deck; the only way to hold one is to make it, live, in a fight. Is the combo the reward for the act of combining, or a skill you can come to own? Options: never lockable, combining is the game; lockable as an earned convenience once both glyphs are highly trained; or lockable but always weaker than a freshly fused copy. Each spends a different part of the system's identity, and the page takes no side.
Seen in: Magicka, where a learned Magick earns a hotkey, the genre's precedent for earned convenience.
Deck-builder eligibility and the no-tree-node status are the two surfaces; interacts with the S17 overlay and with trainer economy if training is the unlock. Priced after the ruling.
When the combo journal ships, does it list every combo from the start, or show silhouettes until each is discovered or taught? Full listing attacks the system's actual problem, invisibility. Mystery makes discovery content, and historically pushes the real list onto fan sites, where it becomes an in-group badge. This page leans full listing with discovery state as flair, and flags it as the taste call that shapes S7 and S11.
Seen in: Chrono Trigger (full listing) versus Magicka (found tomes); Final Fantasy XI is what happens when the game shows nothing.
A flag in the journal design, decided before S7 builds. No code fork until ruled.
Fifty-eight new combos. Every one names the two glyphs that make it, what the fusion does, and why it earns a slot: a most-played pairing served, a silent school seated, a painted icon finally worn, or a dead skill identity revived. Each combo is a proposal of its own, with its own vote arrows and discussion box; argue with them one at a time.
Ground rules that apply to every entry, stated once. Final numbers belong to the combat balance program; these cards describe behavior, never tuning. Control durations do not grow with card count; only damage, buffs, and lingering effects scale, so a deep-stacked combo never means a longer stun. Pet-buffing combos share one channel, so the strongest song or imbue sings alone rather than stacking into something unkillable. Where an effect could misbehave against players, the card says so and the shorter, softer player-versus-player version is part of the proposal. And every pair was checked against the twenty-nine live combos, every pairing promised on a school page, and every other entry here; the slate is collision free. Three independent reviews, one for data, one for balance, one for naming and voice, went over every entry, and their fixes are already folded in.
The ledger of found materials: ten of these combos wear icons that have sat finished and unused in the game files, and two revive dead skill identities reserved years ago. The cheapest entries are pure reassembly of effects the game already knows how to do.
| # | Combo | Glyphs | Schools |
|---|---|---|---|
| S23 | Feral Crescendo | Savage Sonata + Tame Frenzy | Bard + Taming |
| S24 | Lullaby | Mesmerizing Melody + Tame Heal | Bard + Taming |
| S25 | Marked Quarry | Aimed Shot + Tame Taunt Training | Ranged + Taming |
| S26 | Coursing Shot | Disabling Shot + Tame Speed Training | Ranged + Taming |
| S27 | Thunderstroke | Offhand Attack + Lightning | Tactics + Air |
| S28 | Static Bulwark | Fortify Defenses + Shield Of Air | Tactics + Air |
| S29 | Sword Storm | Whirling Blades + Reckless Stance | Blades + Tactics |
| S30 | Counterstroke | Riposte + Defensive Stance | Blades + Tactics |
| S31 | Parting Shot | Escape + Aimed Shot | LightArmor + Ranged |
| S32 | Tumbling Barrage | Tumble + Multishot | LightArmor + Ranged |
| S33 | Storm Fangs | Shock Weapon + Tame Strength Training | Air + Taming |
| S34 | Following Wind | Gust + Tame Speed Training | Air + Taming |
| S35 | Emberhide | Summon Fire Elemental + Summon Tamed | Fire + Taming |
| S36 | Kindled Fury | Ignite Weapon + Tame Frenzy | Fire + Taming |
| S37 | Forked Volley | Chain Lightning + Multishot | Air + Ranged |
| S38 | Windlance | Dash + Piercing Shot | Air + Ranged |
| S39 | Goading Cut | Taunt + Rend | HeavyArmor + Blades |
| S40 | Vanguard Sweep | Body Slam + Whirling Blades | HeavyArmor + Blades |
| S41 | Pulverize | Ground Pound + Earthquake | Bludgeon + Earth |
| S42 | Wyrmfire | Fire Arrow + Rapid Fire | Ranged + Fire |
| S43 | Bedrock Built | Dig In + Earth's Embrace | Shield + Earth |
| S44 | Soul Skewer | Skewer + Tap Soul | Polearms + Death |
| S45 | Eclipse | Solar Flare + Moon Beams | Sun + Moon |
| S46 | Penumbra | Blind + Vanish | Sun + Moon |
| S47 | Phantom Step | Night Step + Blink | Moon + Air |
| S48 | Wildsong | Animal Charm + Chant of Charming | Moon + Bard |
| S49 | Cacophony | Mind Wipe + Atonal Aria | Focus + Bard |
| S50 | Enthrall | Mind Lock + Mesmerizing Melody | Focus + Bard |
| S51 | Illuminated Mind | Spell Binder Stance + Enlightenment | Focus + Sun |
| S52 | Aetheric Ward | Aetheric Feedback + Shield Of Air | Focus + Air |
| S53 | Hymn of Flame | Savage Sonata + Ignite Weapon | Bard + Fire |
| S54 | Cleansing Chorus | Refrain Of Resistance + Purify | Bard + Life |
| S55 | Death Knell | Dirge Of Dissonance + Death Ray | Bard + Death |
| S56 | Withering Tide | Douse + Death Touch | Water + Death |
| S57 | Plague Wind | Death Field + Gust | Death + Air |
| S58 | Sunbolt | Searing Ray + Lightning | Sun + Air |
| S59 | Sleet Storm | Cone Of Cold + Gust | Water + Air |
| S60 | Steam Blast | Fireball + Douse | Fire + Water |
| S61 | Searing Banishment | Banish Undead + Searing Ray | Life + Sun |
| S62 | Vampiric Strike | Sneak Attack + Death Touch | Subterfuge + Death |
| S63 | Deep Slumber | Sap + Torpor | Subterfuge + Earth |
| S64 | Blindside | Opportunity Strike + Dazzling Rays | Subterfuge + Sun |
| S65 | Darkness | Juxtapose + Light | Chaos + Sun |
| S66 | Desiccation | Juxtapose + Rehydration | Chaos + Water |
| S67 | Cold Shoulder | Juxtapose + Taunt | Chaos + HeavyArmor |
| S68 | Lethargy | Juxtapose + Sprint | Chaos + LightArmor |
| S69 | Mirror Beast | Chaotic Clone + Summon Tamed | Chaos + Taming |
| S70 | Smoke and Mirrors | Chaotic Clone + Vanish | Chaos + Moon |
| S71 | Blank Stare | Tabula Rasa + Mind Wipe | Chaos + Focus |
| S72 | Gambler's Stance | Chaotic Feedback + Reckless Stance | Chaos + Tactics |
| S73 | Treachery | Chaotic Step + Sneak Attack | Chaos + Subterfuge |
| S74 | Chaotic Purge | Chaotic Corruption + Purify | Chaos + Life |
| S75 | Prismatic Volley | Chaos Bolt + Multishot | Chaos + Ranged |
| S76 | Warped Rebuke | Chaotic Shield + Deflect | Chaos + Shield |
| S77 | Fiendish Thrall | Summon Daemon + Summon Skeleton | Chaos + Death |
| S78 | Siren Song | Control Demon + Chant of Charming | Chaos + Bard |
| S79 | Twisting Agony | Torpid Torment + Skewer | Chaos + Polearms |
| S80 | Scatterbrain | Confusion + Chain Lightning | Chaos + Air |
The census is the brief: the most-played two-school builds in the game, most with no combo at all. Two combos for each of the biggest pairs, one damage-leaning and one utility-leaning, so every popular build finds at least one that fits how it already plays.
Savage Sonata + Tame Frenzy
The savage melody whips your pet into a howling fury. Your pet gains a stacking attack speed and damage buff for a short while, and a fainter echo of the song grants you a small damage buff for the same duration.
Why this one: Bard+Taming is the #1 build in the game (271 pairs) with zero combos, and both schools have zero combo participation. Builds from: TameFrenzy's own pet buff graph at higher amplitude plus a standard stacking self buff.
Mesmerizing Melody + Tame Heal
A slow, soothing song settles over the fight. Nearby enemies are mesmerized while the melody holds, and your pet is wrapped in a strong heal over time for the duration. Against players the trance is shorter, and any damage breaks it.
Why this one: Second option for the top pair, Bard+Taming 271, covering the control and pet support roles. Builds from: existing mesmerize/charm vocabulary plus the TameHeal pet HoT; the new wiring is running both target classes off one rune (CleansingRain precedent).
Aimed Shot + Tame Taunt Training
Your shot marks the prey and your pet answers. The pet immediately taunts the marked target, and both you and your pet deal bonus damage to it for a short window.
Why this one: Ranged+Taming is the #2 build (207 pairs) with zero combos. Builds from: taunt/threat vocabulary plus a damage-taken debuff (the mark); new wiring is the pet taunt command riding a player attack.
Disabling Shot + Tame Speed Training
Cripple the prey and loose the hound. The shot slows the target's movement, and your pet gains a burst of move speed plus bonus damage against that slowed target while the cripple lasts.
Why this one: Second option for Ranged+Taming 207, the classic hunter and hound fantasy the census says people are already playing. Builds from: existing slow debuff plus the TameSpeedTraining pet speed buff, with a conditional damage rider.
Offhand Attack + Lightning
Both weapons come down as one and thunder bursts between them. The struck target takes weapon damage plus air damage with a brief stun chance, and the clap splashes air damage, and only damage, to enemies around it.
Why this one: Air+Tactics is 123 pairs with zero combos, and Tactics has zero combo participation; this puts OffhandAttack into the combo table for the first time. Builds from: standard attack graph, small AoE burst, and existing stun chance vocabulary.
Fortify Defenses + Shield Of Air
Your defenses knit into a crackling wall of air. You gain a damage absorbing shield for a moderate duration, and melee attackers take a jolt of air damage each time they strike it.
Why this one: Second option for Air+Tactics 123; FortifyDefenses is a zero-combo Tactics active whose only claimed pair is the Tactics page's intra-school Iron Gate. Builds from: existing absorb shield buff plus the Immolation style retaliation damage rider.
Whirling Blades + Reckless Stance
You abandon your guard entirely and become a wheel of edges. A wide whirling attack strikes everything around you for heavy bonus damage, but you take increased damage for a few seconds afterward.
Why this one: Blades+Tactics is 118 pairs with zero combos, and it wears the unused Sword_Storm combo icon as-is. Builds from: WhirlingBlades AoE graph plus RecklessStance's existing damage-up, defense-down stat trade.
Riposte + Defensive Stance
You settle into a patient guard. For a short window, the next melee attack that lands on you is answered automatically with a riposte at bonus damage, and answering it drops a little of your threat.
Why this one: Second option for Blades+Tactics 118, claiming the unused Counter_Attack icon; both glyphs' claimed pairs (Misericorde, Iron Gate) leave this pair free. Builds from: the Riposte counter graph and threat vocabulary; new wiring is the on-hit trigger living in a buff.
Escape + Aimed Shot
You spring backward out of reach and fire in the same motion. The leap repositions you a short distance away and the shot slows the target's movement, buying room to keep the range open.
Why this one: LightArmor+Ranged is 115 pairs with zero combos, and LightArmor has zero combo participation; this is the skirmisher identity in one glyph. Builds from: Escape's reposition and an existing slow debuff; new wiring is movement plus attack in one rune.
Tumble + Multishot
You roll clear and loose a fan of arrows the instant you land. Targets struck at close range have a chance to be knocked back, keeping attackers off you as you reset your footing.
Why this one: Second option for LightArmor+Ranged 115, putting a second LightArmor mobility glyph (Tumble) into the table. Builds from: reposition, the Multishot fan, and existing knockback chance; new wiring is the move-then-fire sequence.
Shock Weapon + Tame Strength Training
Storm power floods your pet's muscles and teeth. The pet gains a damage buff for a moderate duration, and each of its attacks deals bonus air damage with a chance to arc a small spark to one nearby enemy.
Why this one: Air+Taming is 110 pairs with zero combos. Builds from: the ShockWeapon imbue graph and chain damage vocabulary; new wiring is pointing a weapon imbue at the pet instead of the owner.
Gust + Tame Speed Training
A driving wind rises at your backs. You and your pet both gain a burst of move speed for a short duration, and the pet's next attack has a knockback chance as it slams in with the gale behind it.
Why this one: Second option for Air+Taming 110, covering the mobility role for the pair. Builds from: existing speed buffs (TameSpeedTraining already does the pet half) plus Gust's knockback chance as a next-hit rider.
Summon Fire Elemental + Summon Tamed
Your tamed creature answers the call wreathed in living flame. It arrives with a fire imbue: its attacks deal bonus fire damage, and melee attackers take a burn each time they strike it.
Why this one: Fire+Taming is 101 pairs with zero combos; summon-variant combos are proven tech (SummonSkeletonMage, the four elemental flips). Builds from: the summon variant pattern, an attunement damage rider, and the Immolation retaliation burn.
Ignite Weapon + Tame Frenzy
The frenzy takes and the pet's jaws catch fire with it. Your pet gains attack speed for a short duration and every hit it lands applies a burning damage over time to the target.
Why this one: Second option for Fire+Taming 101; IgniteWeapon's fuzzy fence only covers Shield runes, so this pair is clean. Builds from: TameFrenzy's buff plus the standard burn DoT; new wiring is the on-hit burn proc on pet attacks.
Chain Lightning + Multishot
Every arrow in the fan carries a spark. Each target struck takes air damage on top of the arrow, and the sparks chain between struck targets, punishing packed groups hardest. Each enemy can be caught by at most one arc per volley.
Why this one: Air+Ranged is 98 pairs and its two live combos both run through AimedShot; this opens the pair's AoE lane with fresh glyphs. Builds from: the Multishot fan and the ChainLightning chain graph; new wiring is the per-hit chain proc.
Dash + Piercing Shot
You dash forward and your momentum rides the arrow. Fires a piercing shot that punches through everything in a line, and targets it passes through are briefly slowed by the shockwave.
Why this one: Second option for Air+Ranged 98, on glyphs with zero live combos between them. Builds from: Dash's reposition, PiercingShot's line pierce, and an existing slow; new wiring is the dash-then-fire sequence.
Weapon schools crossing into magic and each other, past the handful of pairings 2016 authored. Two of these revive finished work: one wears a dead skill identity that has waited nine years, and two more claim icons that were painted for combos never written.
Taunt + Rend
A shallow, insulting slash that no enemy can ignore. Applies a bleeding damage over time and forces the target to attack you, holding its attention while the wound works.
Why this one: Blades+HeavyArmor is 97 pairs with no combo, and this gives sword-and-board tanks a threat opener in one glyph. Builds from: Rend's bleed DoT and Taunt's forced-target threat graph, both live today.
Body Slam + Whirling Blades
You crash into the enemy line and spin out of the impact. The slammed target has a knockdown chance, and everything around you takes a whirling arc of weapon damage as you carve room to stand.
Why this one: Second option for Blades+HeavyArmor 97, covering the frontline breach role; BodySlam's only live pair is Knockback so this stays clean. Builds from: knockdown chance, the WhirlingBlades AoE, and a standard short resist buff.
Ground Pound + Earthquake
You bring the maul down hard enough to rupture the ground itself. Enemies around you take crushing damage with a knockdown chance, and everything struck picks up a stacking armor shred debuff.
Why this one: Revives dead SkillID 1004 under its own unused Pulverize icon, and Bludgeons+Earth is 83 pairs (#12). Not a lingering ground hazard, so it stays clear of the Earth G7 fuzzy fence. Builds from: AoE burst, knockdown chance, and the BreakArmor style stacking shred.
Fire Arrow + Rapid Fire
You loose burning arrows so fast the stream becomes a cone of flame. Channels rapid fire damage across everything in a short cone, and repeated hits stack a burning damage over time on each target, up to a modest cap of stacks.
Why this one: Weapon-school cross pollination for Ranged x Fire, wearing the unused Dragons_Breath icon; RapidFire has zero live combos today. Builds from: the RapidFire channel, a ConeOfCold style cone, and the stacking burn DoT.
Built August 15. Bedrock is in the game as written: combining Dig In with Earth's Embrace in hand makes it, casting needs a shield, and for a short duration you take greatly reduced damage and cannot be knocked back or knocked down while held in place. A cleanse cannot strip the hold, and the resistance and the hold always end together. New icon in both icon sets, the shield-plant animation, and a layered shield and earth cast sound. The first combo of the fifty-eight to land. Not on the test shard yet.
Dig In + Earth's Embrace
You plant the shield and the earth rises to hold you. For a short duration you gain heavy damage resistance and cannot be knocked back or knocked down, but you are rooted in place until it ends.
Why this one: Shield is the least-played school (79 specialists) and both its live combos run through ShieldBash; this gives DigIn a lane and stays off the Purify+Earth fence. Builds from: resist buff, CC immunity flags, and the Root effect turned on the caster.
Skewer + Tap Soul
You run the target through and drain its life back along the shaft. Deals weapon damage plus death damage, and a share of the death damage returns to you as health.
Why this one: Polearms cross pollination beyond its two Water combos, wearing the unused Vampiric_Strike icon; Death has 255 specialists and TapSoul has zero live combos. Builds from: the Skewer attack graph and TapSoul's existing life drain.
Moon, Focus, and Bard have never combined with anything; Subterfuge has one glyph at the table. This group gives every silent school several seats, pairs Sun with Moon at last, and fills the elemental gaps the live table skipped.
Solar Flare + Moon Beams
A ray of blotted sunlight that strikes with sun and moon at once, half of each. Struck targets are briefly blinded, and the weaker of their two attunements is lowered for a short while.
Why this one: Moon has zero combo participation across 12 actives, and Sun (330 specialists) and Moon are already framed as opposites by their warding passives; an eclipse pair is the natural flagship. Reuses: split damage, blind chance, a single non-stacking attunement debuff.
Blind + Vanish
The caster snuffs the light around them, blinding nearby enemies while slipping into stealth. Affected enemies also lose some of their ire toward the caster, letting them slip away unseen.
Why this one: Second eclipse-flavored Sun+Moon pair, defensive twin to Eclipse; Moon zero participation, Sun 330 specialists. Reuses: AoE blind, stealth apply, threat reduction, all existing graphs stapled together.
Night Step + Blink
A single step through shadow that carries as far as the longer of its two parents, leaving the caster briefly stealthed on arrival and shedding threat as they vanish. An escape, not a charge.
Why this one: Moon zero participation; Air is the most combo-friendly mobility school and features in three top-15 spec pairs (Air+Tactics 123, Air+Taming 110, Air+Ranged 98). Reuses: teleport, stealth apply, threat reduction.
Animal Charm + Chant of Charming
A haunting melody that charms a beast harder and longer than either glyph and can take hold on stronger creatures. While charmed, the beast fights for the caster with a small damage bonus.
Why this one: Two zero-participation schools (Moon, Bard) sharing the charm vocabulary; Bard's list is the largest with no combos at all. Reuses: charm, duration rider, charmed-ally combat flag.
Mind Wipe + Atonal Aria
A discordant blast that deals magic damage around the bard and drains focus from every enemy it touches, up to a cap per target. A capped share of the drained focus returns to the caster, and the drain is reduced against players.
Why this one: Focus has zero combo participation despite 261 specialists; Bard also sits at zero. Reuses: burst damage around the caster, focus drain, focus refund.
Mind Lock + Mesmerizing Melody
Locks a single enemy in a deep trance that lasts noticeably longer than a normal mesmerize and halts their focus regeneration while it holds. Any damage breaks the trance.
Why this one: Focus zero participation (MindLock has no combo anywhere); Bard zero. Reuses: mesmerize graph plus a regen debuff rider, both existing vocabulary.
Spell Binder Stance + Enlightenment
A stance that sharpens the caster's mind: spell damage and critical chance rise while it is held. Movement slows slightly while the stance is kept.
Why this one: Focus zero participation; Enlightenment already combines elsewhere (ImbecileIncantation, SummonSkeletonMage) so the pair rides known ground; Sun 330 specialists. Reuses: stance stat package wiring.
Aetheric Feedback + Shield Of Air
A crackling shield that returns a portion of incoming spell damage to the attacker as air damage. Each spell that strikes the ward also refunds a little focus to the caster.
Why this one: AethericFeedback is a zero-combo Focus active in a zero-participation school. Reuses: damage shield and reflect graphs; the on-spell-hit focus refund trigger is the one new piece of wiring.
Savage Sonata + Ignite Weapon
A war song that wreathes the whole party's weapons in flame while it plays, adding fire damage to their attacks. The shared flame burns fainter than a personally ignited weapon, and a personal Ignite Weapon takes its place while it lasts.
Why this one: IgniteWeapon is one of Fire's combo-dead runes and only its Shield pairing is claimed (Fire S10); Bard zero participation. Reuses: weapon imbue effect delivered through the existing party song aura; the party-wide imbue is the new wiring.
Refrain Of Resistance + Purify
A steady chorus that periodically strips one harmful effect from each party member in range and raises their resistances while it plays. The cleanse pulses on the song's existing beat.
Why this one: Bard zero participation; Purify has no combo today and the Life page claims only its Earth pairing (Stoneward Blessing), leaving this open. Reuses: cleanse, resistance aura, song pulse cadence.
Dirge Of Dissonance + Death Ray
The dirge deepens into a funeral toll that deals death damage each pulse to enemies near the bard. Enemies that die under the knell restore a small amount of focus to the bard.
Why this one: Death's three live combos all run through Death Ray or Summon Skeleton; the rest of the tree, the dirge's partner included, is unused. Bard sits at zero. Reuses: aura pulse and drain vocabulary; the on-kill focus return is the new wiring.
Douse + Death Touch
Death curdles the water in the target's veins. Deals death damage over time and slows them as their body stiffens, and the withering flesh loses some fire resistance while the effect runs.
Why this one: No Water+Death combo exists; drawn for the unused Desiccation combo icon; Death 255 and Water 127 specialists. Reuses: DoT, slow, resistance debuff.
Death Field + Gust
Hurls the death field downwind, placing a pestilent cloud at a targeted spot instead of at the caster's feet, with a wider spread. The cloud drifts slowly on the wind, and its ticks are gentler than a field cast at your own feet.
Why this one: No Death+Air combo exists; wears the unused Cloud_of_Pestilence icon; distinct from Earth G7's claimed Grave Soil (no Earth rune, air-thrown cloud not a ground hazard). Reuses: field plus DoT; ranged placement is the new wiring.
Searing Ray + Lightning
A searing bolt of skyfire that strikes instantly, dealing mixed sun and air damage. It carries a chance to blind and a chance to briefly interrupt the target's casting.
Why this one: No Sun+Air combo exists anywhere in the live 29; Sun is the most specialized school at 330. Reuses: bolt damage, blind chance, interrupt chance.
Cone Of Cold + Gust
A howling cone of wind-driven sleet that deals water damage and heavily slows everything caught in it. Targets in the cone have a chance to lose their footing and be knocked down.
Why this one: No Water+Air combo exists; Water 127 specialists and ConeOfCold has no combo despite Water's otherwise heavy combo use. Reuses: cone damage, slow, knockdown chance.
Fireball + Douse
A point-blank eruption of scalding steam that deals mixed fire and water damage around the caster. The mist lingers a moment, lowering the hit chance of enemies caught in the burst.
Why this one: The only live fire-meets-water flavor routes through Chaos (Evaporate is Douse+Juxtapose); this gives the direct pair a combo and wears the unused Steam_Blast icon. Reuses: PBAoE damage, hit-chance debuff.
Banish Undead + Searing Ray
A column of banishing light that deals heavy bonus damage to undead and blinds them briefly. Against the living it lands as a lesser flash of searing light.
Why this one: Life sits in only two live combos and the Life page's Sun claim covers HealingRay only (Radiant Dawn), leaving BanishUndead's Sun pairing open; Sun 330 specialists. Reuses: vs-type bonus damage, blind, small AoE heal.
Sneak Attack + Death Touch
A stealth opener that drains life from the victim, returning a modest fixed share of the damage as healing to the attacker. Landed from stealth it also applies a short death damage over time effect.
Why this one: Subterfuge participates only through Trip and SneakAttack is unused; wears the unused Vampiric_Strike icon; Death 255 specialists. Reuses: stealth-conditional opener, TapSoul-style life drain, DoT.
Sap + Torpor
Puts the target into a heavy sleep that lasts longer than a standard mesmerize; any damage wakes them. On waking, the target remains slowed for a few seconds as the torpor clings.
Why this one: Sap has zero combos (Subterfuge is Trip-only) and Torpor has zero despite Earth's popularity as a combo school; clears the Earth G7 fuzzy claim because it is a single-target sleep, not a ground hazard. Reuses: mesmerize and slow; the on-wake rider is the new wiring.
Opportunity Strike + Dazzling Rays
An opportunistic strike that deals bonus damage to targets that are blinded, stunned, rooted, or knocked down. Against any other target it lands as a normal hit but applies a brief blind, setting up the next one.
Why this one: OpportunityStrike is an unused Subterfuge active in a one-rune school; Sun 330 specialists; the conditional bonus logic already lives in OpportunityStrike itself. Reuses: conditional bonus damage, blind apply.
Chaos is the transmutation school, and today it combines almost entirely through one glyph. Sixteen combos put the whole chaos kit to work across the rest of the game: a few new Juxtapose inversions in the school's signature grammar, and thirteen combos for the chaos glyphs that have never combined with anything.
Juxtapose + Light
Inverts the light spell into a patch of unnatural gloom at the target point. Enemies inside are blinded and suffer reduced accuracy, while allies inside gain a brief concealment bonus. The gloom fades after a short duration.
Why this one: The signature Juxtapose inversion the icon set has been waiting for: the unused Darkness combo icon exists. Sun (330 specialists) has one damage combo, Rainbow Blast; its only other participation is Enlightenment's two utility pairs. Reuses ground field, blind, and concealment vocabulary.
Juxtapose + Rehydration
Inverts the hydrating stream and wrings the moisture out of the target instead. Deals a lingering drying damage over time and saps a small amount of focus with each tick as the body withers.
Why this one: A heal inverted, the exact Juxtapose grammar the territory calls for, and it wears the unused Desiccation combo icon under its own name (the Pulverize precedent). Water sits at 127 specialists. Reuses DoT plus focus-drain effect vocabulary only.
Juxtapose + Taunt
Inverts the taunt into pure dismissal. You shed a large portion of your threat against nearby enemies and are treated as a far less attractive target for a short time.
Why this one: A Juxtapose inversion of a tank verb gives every non-tank a threat dump. Taunt has zero combo use, and Heavy Armor's only live combo is Titan Smash. Pure reuse of the existing taunt and threat vocabulary, sign flipped.
Juxtapose + Sprint
Inverts the burst of speed and pours it onto an enemy as leaden weight. The target suffers a strong movement slow and a milder attack speed slow for a moderate duration.
Why this one: Light Armor has zero combo participation despite LightArmor+Ranged being the 7th most popular spec pair (115). Sits in the Clumsy and Enfeeblement family of quirky Juxtapose debuff names. Pure reuse of existing slow-debuff vocabulary.
Chaotic Clone + Summon Tamed
Calls a shimmering illusion of your active pet to fight beside the original for a short time. The copy deals no true harm; it taunts and harries, and detonates into a burst of confusion on nearby enemies when it expires.
Why this one: Taming has zero combo participation and anchors the two biggest spec pairs (Bard+Taming 271, Ranged+Taming 207); this is ChaoticClone's first cross-school use. New wiring only: summon-variant spawn of the active pet archetype at reduced stats plus the existing confusion burst.
Chaotic Clone + Vanish
You slip into hiding and a clone appears exactly where you stood. The clone taunts nearby enemies for a few seconds, holding their attention while you reposition or escape.
Why this one: Moon has zero combo participation (120 specialists), and this pairs the two most theatrical escape glyphs in the game. Pure effect-graph reuse: stealth entry, the existing clone summon, and the taunt rider.
Tabula Rasa + Mind Wipe
Wipes both minds at once. You discard and redraw your hand, and the target loses focus for each of the first three glyphs you redraw, standing briefly confused while its thoughts reassemble.
Why this one: Focus has zero combo participation despite 261 specialists, TabulaRasa appears in no combo, and the two glyphs are flavor twins (both erase). Deck manipulation with teeth per the chaos identity brief. New wiring: TabulaRasa redraw feeding a per-card focus-drain count plus existing confusion.
Chaotic Feedback + Reckless Stance
A gambler's stance. While it is held, every damaging or healing glyph you play has a chance to be amplified and a smaller chance to backfire for minor damage to you, on top of a modest boost to damage dealt. Control effects and summons are never amplified.
Why this one: Tactics has zero combo participation while carrying two top-10 spec pairs (Air+Tactics 123, Blades+Tactics 118), and ChaoticFeedback has zero use outside the claimed Wild Surge. The gambling-payoff identity in stance form. New mechanism: a while-stance-active per-glyph proc hook.
Chaotic Step + Sneak Attack
You blink directly behind your target and drive the blade home in the same motion. The strike counts as attacking from behind and gains bonus critical chance.
Why this one: Subterfuge participates through Trip alone (135 specialists), ChaoticStep has no cross-school combo, and the unused Backstab combo icon is a made-to-order fit. Pure reuse: existing teleport plus the positional sneak-attack bonus.
Chaotic Corruption + Purify
Scours every affliction from your body and feeds it to the chaos. Each debuff removed hurls a bolt of a random damage type at a random nearby enemy.
Why this one: Revives dead SkillID 1006 with exactly its named flavor; Purify+chaos is explicitly free while Purify+Earth belongs to Life S11. A side-effect bargain in the chaos register. New wiring: existing cleanse counting removed debuffs into chaos-bolt procs.
Chaos Bolt + Multishot
Fires a fan of arrows, each one transmuted into a different random damage type in flight. Each arrow carries a small chance of a matching minor rider such as a brief burn, chill, or shock.
Why this one: Ranged anchors three top-10 spec pairs (Ranged+Taming 207, LightArmor+Ranged 115, Air+Ranged 98) yet chaos never touches it; second home for zero-use ChaosBolt. New wiring: per-projectile random-type roll on the existing multishot graph plus existing minor riders.
Chaotic Shield + Deflect
Raises a warped guard for a short time. While it holds, a portion of the damage you block is thrown back at the attacker as a randomly chosen damage type.
Why this one: Shields is the least popular school in the game (79 specialists, last place) and needs a marquee toy; ChaoticShield appears in no combo; the unused Counter_Attack icon fits a reflect exactly. New wiring: block buff plus damage-return with the random-type roll.
Summon Daemon + Summon Skeleton
Binds a daemon into dead bones, raising a horned skeletal brute that hits harder than either summon alone. When it lands a killing blow it briefly frenzies and hurls itself at the nearest enemy, heedless of orders.
Why this one: SummonDemon appears in no combo, Death has 255 specialists, and the summon-variant pattern (SummonSkeletonMage precedent) makes this the cheapest possible chaos summon: a variant archetype plus the existing frenzy behavior, with a chaos downside baked in.
Control Demon + Chant of Charming
A charm that no longer cares what it is charming. Briefly seizes a single creature of almost any type, undead and daemons included, and turns it against its own allies. The mightiest creatures shrug it off, other players' pets cannot be taken, and victims come out of it confused.
Why this one: ControlDemon appears in no combo and Bard is the zero-participation school behind the number one spec pair (Bard+Taming 271), so a second bard entry pushes the biggest gap. New wiring: existing charm graph with a widened creature-type mask plus confusion on expiry.
Torpid Torment + Skewer
Impales the target and leaves the torment twisting in the wound. Applies a stacking damage over time that slows the victim more with each stack.
Why this one: TorpidTorment appears in no combo and this gives polearms a chaos entry without touching the Puncture+Skewer pair Polearms S22 claimed. Pure reuse: hit, stacking DoT, scaling slow, all existing debuff vocabulary.
Confusion + Chain Lightning
Lightning that argues with itself. Arcs between nearby enemies as normal chain lightning, and every enemy struck is briefly confused, wandering off target as the charge scrambles its wits.
Why this one: ChaoticConfusion's only combo is intra-chaos (ChaosBomb); this is its first cross-school use, aimed at the large Air population (Air+Tactics 123, Air+Taming 110, Air+Ranged 98). Pure reuse: existing chain-lightning graph with the existing confusion debuff on each jump.
Fifteen of the proposals above touch what a player sees, and the anti-pattern would be fifteen widgets. They compose into one surface plus three whispers. The surface is the Combo Codex: the existing combo list, moved where players look, made truthful, and grown into a journal. The whispers are the shimmer in the hand, the breadcrumb in every glyph tooltip, and the name in the combat log. Everything else on this page is data, tuning, or content that feeds them.
| Combo | Made from | Level | |
|---|---|---|---|
| ★ | Flaming Bash | Shield Bash + Flame Fist | 74 |
| ★ | Fiery Decay | Fire Arrow + Death Ray | 61 |
| ☆ | Shock Arrow | Aimed Shot + Discharge | 58 |
| ☆ | Sacrifice | Healing Ray + Death Ray | — |
| ☆ | Titan Smash | Knockback + Body Slam | — |
Five phases, ordered so that no phase points players at something the previous phase has not made true. The telemetry clock starts in phase one on purpose: by the time the retune arrives in phase five, the data is thirty days deep.
| Phase | Theme | Cards | Gate |
|---|---|---|---|
| 1 | Make the table true, start the clock | S2 S3 S4 S5 S15 | none; all small |
| 2 | Point at it | S1 S6 S9 S10 | phase 1 hygiene |
| 3 | Celebrate and record | S22 ruling, then S7 S8 S12 | S6 home for the journal |
| 4 | Teach in the world, deepen the machine | S11 S13 S14 S17, then the compendium slate | S7 journal; S14 before new combos |
| 5 | Retune and the frontier | S16, S19 pilot; rule on S18 S20 S21 | 30 days of clean S15 data |
flowchart TD
subgraph P1[Phase 1: true table]
S2[S2 labels] --> S6[S6 list anywhere]
S5[S5 filter] --> S6
S4[S4 lint]
S15[S15 telemetry]
end
subgraph P2[Phase 2: point at it]
S6 --> S9[S9 tooltip breadcrumbs]
S1[S1 popup] ~~~ S10[S10 shimmer]
end
subgraph P3[Phase 3: celebrate]
S22[S22 ruling] --> S7[S7 journal + fanfare]
S6 --> S7
S7 --> S8[S8 announcements]
S7 --> S12[S12 achievements]
end
subgraph P4[Phase 4: teach and deepen]
S7 --> S11[S11 trainers and tomes]
S7 --> S13[S13 practice]
S14[S14 alternate pairings] --> CMP[Compendium slate]
end
subgraph P5[Phase 5: retune]
S15 --> S16[S16 stack math]
S15 --> S19[S19 status pilot]
S14 --> S18[S18 higher order]
end
TS20[Tactics page S20 mechanism] -.-> CMP
Cross-page honesty: the melee sword-form combos in the compendium ride the combo mechanism card the Tactics page owns, exactly as the Shield and Light Armor pages already queue behind it. The retune (S16) shares the thirty-day clean-telemetry gate the Ranged page committed to after its combo reporting fix. Nothing else on this page blocks on anyone.
A combo is data, not engine work: a new entry in the skill list, a result glyph authored from an existing effect vocabulary, a pairing line naming its two glyphs, a name and description, and an icon. Thirty-one finished combo icons already sit unused in the game files, and several proposed combos below claim one. The expensive part of a combo is deciding what it does; the wiring is a day, not a month. That economy is what makes a fifty-combo compendium a realistic proposal rather than a wish.
The combo system already carries editor test coverage this page extends rather than invents: the combo list's row layout and sorting, the elemental summons' attunement recursion guard, the combo identity round-trip in use reporting, charged-glyph card counts, and the R key's default binding all have pinned tests. The Tier 1 cards each add their own pin: label list equals combo table, lint rules, filter semantics, safe defaults, and, with S14, the or-rendering of alternate pairings.
| Claim | Source |
|---|---|
| 29 combos and their glyphs | the live combo data, read directly from the game files |
| Which schools sit at zero | the same combo table, cross-checked against every school's skill list |
| Most-played school pairings | specialization census of 120,337 characters, August 2026 |
| Combo usage counts on school pages | 30-day skill-use telemetry, quoted from the sibling school pages |
| Player asks | forum threads 2016 to 2026: combo visibility, charge times and stacking, glyph keybinds |
| Genre mechanics | Final Fantasy XI, Chrono Trigger, Guild Wars 2, Magicka, Divinity Original Sin 2, Elder Scrolls Online, Path of Exile, Outward, Lost Ark, Street Fighter II, each verified against its own documentation |
Every proposal card and every compendium combo above has its own vote arrows and discussion box, and the general thread below takes everything else. The combos are deliberately written to be argued with: if a combo's effect reads wrong for its school, or the pairing you actually play is missing, that is exactly what the comment box is for.
No account needed. Every proposal above also has its own Discussion box for talking about that one alone; this thread is for the plan as a whole. Comments are plain text, held to basic decency, and may be trimmed.